The Older Workers and Retirement Chartbook Chapter 3. Risk

What economic risks do older Americans face?

Americans face increasing economic risks as they age, including risks associated with poor health, job loss, and financial market downturns. Aging increases the risk of developing health conditions that are expensive to treat, affect a person’s ability to earn a living, or result in the need for assistance with daily activities. Older workers are more likely than younger workers to assume caregiving responsibilities for aging family members that interfere with work. And older workers who lose their jobs are at greater risk of significant earnings losses than younger workers because they are likely to be unemployed longer, to accept a job with lower pay, or to take unplanned early retirement. The COVID-19 pandemic and the ensuing recession exacerbated many of these risks, though economic impact payments, expanded unemployment insurance benefits, and other temporary measures enacted by Congress helped cushion the financial blow.

The Older Workers and Retirement Chartbook

A joint project of EPI and the Schwartz Center for Economic Policy Analysis

Older households, especially those relying on 401(k) plans to fund retirement, are also exposed to risks associated with asset price volatility. Though older households may experience smaller percentage declines in net worth than younger households when stock and housing markets collapse, their losses are larger in dollar terms and have more of an impact on their future standard of living, which depends more on accrued assets and less on earnings than that of younger households (authors’ analysis of Federal Reserve 2022).

Jump to the charts

Social insurance programs help shield older workers and retirees from economic risks associated with aging. Social Security provides inflation-adjusted lifetime benefits to retired and disabled workers, spouses, and survivors. Supplemental Security Income (SSI) and other means-tested programs help support low-income disabled and elderly Americans. Medicare provides health insurance to most Americans at age 65 and to some disabled workers, while Medicaid provides health insurance to many lower-income Americans and covers the cost of long-term care for those who have exhausted other resources. Workers’ compensation pays for medical care and provides cash benefits for workers who are injured on the job or contract work-related illnesses, and unemployment insurance replaces a share of lost earnings for workers who are laid off. (This is not an exhaustive list.)

These programs offer critical protections, but they need to be strengthened. They can be difficult to access and often provide less generous benefits or less comprehensive coverage than programs in peer countries, mainly because the United States has historically relied on employers to offer retirement, disability, health, and other critical benefits (NASI, forthcoming; Hacker 2002, 2006; OECD 2020). Relying on employers to provide essential benefits leaves many workers uncovered and can cause employers to avoid hiring those who would be more expensive to insure.

The limitations of social insurance programs are the outcome of an age-old political tug-of-war over the proper role of government in the United States, not necessarily of problems inherent to social insurance (Hacker 2002 and 2006; Dionne 1999). As demonstrated in countries with stronger social protections, government-provided or government-mandated social insurance has many potential advantages over voluntary private insurance, though whether these advantages are realized depends on public support and the government’s administrative capabilities.

An obvious advantage of social insurance programs is that overhead costs and risks can be spread over larger groups and, in many cases, over citizens’ lifetimes and across generations. Universal coverage also avoids the problem of adverse selection—the fact that people at higher risk are more likely to buy insurance. Adverse selection drives up costs and can make insuring against some risks, such as the need for long-term-care, unaffordable to most people (Sammon 2020).

Other potential advantages of government-sponsored social insurance programs may be less well known. Access to social insurance can encourage people to adopt behaviors that have broader societal benefits, such as getting vaccinated, or to take entrepreneurial risks in starting a business, knowing there is a safety net for them and their families (Frick 2015). Many forms of social insurance, including Social Security and unemployment insurance, help stabilize the economy by expanding government spending during recessions and contracting spending during recoveries (Ghilarducci, Saad-Lessler and Fisher 2011; Dolls, Fuest, and Peichl 2010). Government programs are also better equipped to withstand systemic risks such as demographic changes, financial crises, or pandemics, which can drive private companies into bankruptcy.

Social insurance programs can ease portability, lower administrative costs, and reduce the time users spend navigating the system. For example, health care users in Canada and other countries with single-payer systems have less paperwork to contend with than their U.S. counterparts, who need to be vigilant about whether providers, procedures, and prescriptions are covered by their insurance and often need to challenge denied claims. Similarly, U.S. health care providers must hire staff to negotiate with and bill multiple insurers (Santhanam 2020; Frakt 2019). The Congressional Budget Office has estimated that a single-payer system could save over half a trillion dollars in administrative costs annually, or 1.8% of GDP (Nelson 2022). In addition to being more efficient, social insurance takes the profit motive out of coverage and claims decisions and permits a more egalitarian allocation of health care resources than if they are apportioned based on ability to pay.

While Medicare and Medicaid offer some of the efficiency advantages of single-payer programs (CBO 2022a), these programs are not available to everyone. Among social insurance programs, universal or quasi-universal programs such as Medicare reduce administrative costs and barriers to access compared with means-tested programs such as Medicaid. However, determining eligibility is an inherent function of disability, workers’ compensation, and some other social insurance programs.

No one disputes that consumer interest and competition have led to higher quality and lower prices for many goods and services, from smartphones to streaming media. But in sectors where knowledge and trust loom large and market power is concentrated, such as health care, consumer involvement has proven relatively ineffective at lowering costs, improving quality, or spurring technological advances compared with government action (CBO 2022b; Cleary, Jackson and Ledley 2020). In such sectors, the size and expertise of government agencies can be leveraged to negotiate better-quality and cheaper products and services than what inexpert and dispersed individual consumers can negotiate. Thus, for example, few dispute that the Thrift Savings Plan, a retirement plan for federal employees, offers more cost-effective and appropriate investment options than what most retirement savers have in their 401(k) accounts, which is why it has become a model for bipartisan reform proposals (Lunney 2016; Ghilarducci and Hassett 2021).

Single-payer health care and other features of a Canadian- or European-style social insurance system may not be politically realistic in the United States in the near future, but there should be room to expand efficient programs that already have strong support among voters. Of course, all this depends on effective governance. The potential advantages of social insurance may not materialize if government actors are overly bureaucratic, corrupt, or pay too much or too little attention to costs.

What do the charts tell us?

The American patchwork system of private insurance and social insurance leaves many without coverage and results in an inequitable distribution of care and costs. Our research finds that 1 in 10 older households pays $13,800 or more in annual out-of-pocket medical expenses (Chart 3A). Nearly 3 in 10 seniors age 65 or older who live in poverty would not be poor absent medical expenses (Chart 3B). Black and Hispanic households—already at greater risk of poverty in old age—are more likely than white households to fall into poverty because of medical expenses (Chart 3C).

Older Americans needing long-term care at home or in institutional settings bear some of the highest costs. Most who need long-term services and supports for two or more years end up on the means-tested Medicaid program after exhausting most of their savings (Chart 3D). Though the lowest-income seniors are at greatest risk of needing long-term care, risks are elevated across the income spectrum because higher-income Americans, who tend to be healthier at younger ages, also tend to live longer and to develop cognitive impairments and other conditions associated with advanced old age (Chart 3E).

While most American households strengthened their ability to weather financial shocks in the recovery that followed the Great Recession of 2008–2009 (Federal Reserve Board of Governors 2022), older lower-income, Black, and Hispanic households saw an increase in financial fragility (Charts 3F and 3G). One factor contributing to the financial fragility of these older households was an increase in education debt (Chart 3H).

Thus, older households were ill-prepared for the pandemic recession in 2020, in which—unusually—older workers were more likely to lose their jobs (Chart 3I). These job losses persisted longer than for younger workers and employment at older ages remains depressed, especially among those age 65 and older and non-college-educated workers (Charts 3J and 3K).

What can we do to better insure older Americans against financial risks?

As discussed earlier, Social Security retirement benefits replace a shrinking share of earnings even as fewer workers are covered by secure pensions. Expanding Social Security benefits is the most important action we can take to boost retirement security while strengthening older workers’ bargaining power and reducing their reliance on earnings from low-paid jobs.

Other top priorities are expanding benefits for low-income older adults and making it quicker and easier for workers with disabilities to access benefits. Many low-income disabled and elderly adults are unable to meet SSI’s stringent and outdated eligibility requirements, yet SSI benefits are so meager that half of beneficiaries live in poverty (Nuñez 2022; CBPP 2022). Eligibility rules should be changed and benefits increased as proposed in the Supplemental Security Income (SSI) Restoration Act (Romig 2021).

Disabled workers applying for Social Security Disability Insurance (SSDI) benefits must demonstrate that they have severe impairments that preclude them from engaging in “substantial gainful activity” anywhere in the country—generally, the ability to earn $1,350 per month in 2022—regardless of whether such work is realistically available to them. The time-consuming and complicated process of accessing benefits discourages many from applying, a problem that disproportionately affects less educated workers and has gotten worse with budget cuts, especially during the pandemic (CBPP 2021; Deshpande and Li 2019; Weaver 2021; Romig 2022). Delays caused by a lengthy application process and administrative backlogs are compounded by a five-month waiting period for SSDI benefits to start and an additional two-year waiting period, after applicants begin receiving SSDI benefits, for Medicare coverage. Many applicants in poor health get sicker without health coverage and do not live long enough to receive benefits (Nin 2015). SSDI should receive more administrative funding, and Congress should take steps to ensure timely and equitable access to benefits, including eliminating unnecessary waiting periods (Romig 2021; Gronniger 2022;  Lilly 2022; Vallas 2022).

Medicare and Medicaid provide affordable health care coverage for many older workers and retirees. These programs are especially important for older women of color, who confront multiple and compounding forms of discrimination over their lifetime, with devastating impacts on their health and economic security at older ages (NWLC & Justice in Aging 2021).  The Affordable Care Act of 2010 and the Families First Coronavirus Response Act of 2020 expanded Medicaid eligibility to more low- and moderate-income Americans, and the Inflation Reduction Act of 2022 took steps to limit drug costs for Medicare beneficiaries, including by allowing the federal government to negotiate directly with providers for some drug prices.

We should defend and build on these successes. Some state lawmakers have resisted Medicaid expansion under the Affordable Care Act, the pandemic health emergency expansion is set to expire, and some Republicans have vowed to roll back the Inflation Reduction Act (Cubanski, Neuman, and Freed 2022; Choi 2022; Wagner and Erzouki 2022). The Medicaid coverage gap currently leaves 2.2 million low-income people in 12 states without coverage, and most people are not eligible for Medicare until age 65 (Ammula and Rudowitz 2022). And because Medicare does not limit out-of-pocket expenses, beneficiaries must purchase supplemental coverage or they might find themselves having to choose between going into medical debt or forgoing expensive and needed care (Kaiser Family Foundation 2019; Madden et al. 2021).

Lowering the Medicare eligibility age and closing the Medicaid coverage gap would expand affordable coverage, increasing health care utilization and reducing health care costs by offering access to preventative and screening services and reducing the incidence of conditions caused or worsened by postponing treatment (Garfield, Rae, and Rudowitz 2021; Kilbourne 2005).

Medicaid is not a substitute for long-term care insurance since it is only available to people who have drawn down other resources or had low incomes and savings to begin with. Public long-term care insurance schemes such as Washington state’s new WA Cares Fund allow costs to be spread across an individual’s working life and pool the risk across generations (Veghte 2021). Reforms should also make it easier for people to receive home- and community-based services and should provide more support to family caregivers so that people can stay in their homes as they age or when they need long-term care for other reasons.

Last, we should address employment risks faced by older workers. Many unemployed workers are not covered by unemployment insurance or receive inadequate benefits—a problem due in part to wide variations in eligibility rules and benefits across states (Bivens et al. 2021). Similarly, cutbacks to workers’ compensation programs in many states have had severe consequences for many workers who are injured or fall ill on the job (Grabell 2015).


Older Americans face the risk of high medical costs—even when covered by Medicare: Medical expenses by age and expenditure percentile, 2021

3A
Percentile Ages 55–64 Age 65+
50th $3,900 $4,900
75th $8,200 $8,800
90th $13,800 $13,800
95th $18,900 $17,600
99th $32,800 $29,800
ChartData Download data

The data below can be saved or copied directly into Excel.

Notes: Households are ranked by the amount they spent on health insurance premiums and out-of-pocket medical expenses in 2021. The 50th percentile, or median, is the amount the typical household spends (50% spend less and 50% spend more). Medical expenses are based on estimates used for the Supplemental Poverty Measure (SPM), an alternative poverty measure published by the Census Bureau since 2010. This measure includes health insurance premiums, co-pays, prescriptions, medical supplies, and over-the-counter expenditures such as vitamins and pain relievers (Creamer 2022).

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of IPUMS Current Population Survey microdata (Flood et al. 2021).

Americans of all ages, including older Americans, are underinsured against medical expenses, including premiums, co-pays, and other out-of-pocket costs. Most Americans are eligible for Medicare at age 65, leaving only 1.2% in the 65+ age group uninsured in 2021 (Keisler-Starkey and Bunch 2022). But Medicare coverage is no guarantee against high out-of-pocket costs. A typical Medicare-eligible senior still spends nearly $5,000 on health care in a year; 10% spend $13,800 or more; and the unluckiest 1% spend $29,800 or more.

Older Americans ages 55–64 have similarly high out-of-pocket costs. The average amount this group spends is $6,100, versus $6,600 for Medicare-eligible seniors (averages not shown in chart). Those with expensive conditions actually spend more than seniors—$18,900 or more for the unluckiest 10%, versus $17,600 or more for their 65 and older counterparts.

Older Americans ages 55–64 are healthier on average than those age 65 and older but are less likely to have government-provided health insurance. Most do not qualify for Medicare, and while the expansion of Medicaid eligibility under the Affordable Care Act reduced uninsured rates among lower-income older Americans, 9.4% of Americans ages 55–64 remained uninsured in 2021 (Katch, Wagner, and Aron-Dine 2018; Keisler-Starkey and Bunch 2022; authors’ analysis of IPUMS Current Population Survey microdata [Flood et al. 2021]). Though most out-of-pocket costs are capped for those with insurance (Rae, Amin, and Cox 2022), some still face high costs.

Share this chart:

Copy the code below to embed this chart on your website.

Among seniors in poverty, nearly 3 in 10 have been driven into poverty by medical expenses: Poverty rates by age, with and without medical expenses, 2021

3B
Supplemental Poverty Measure (SPM) poverty rates by age group, 2021 Poverty with or without medical expenses Poverty only with medical expenses Total poverty rate
Ages 25–54 6.6% 1.2% 0
Ages 55–64 7.0% 1.8% 0
Age 65+ 6.8% 2.7% 0
ChartData Download data

The data below can be saved or copied directly into Excel.

Notes: “Poverty only with medical expenses” is the share whose medical expenses push them into poverty (who would not otherwise be in poverty). “Poverty with or without medical expenses” is the share who are in poverty even without their medical expenses. Medical expenses are based on estimates of health insurance premiums and out-of-pocket health costs used to estimate the Supplemental Poverty Measure (SPM). The SPM is an alternative poverty measure published by the U.S. Census Bureau since 2010.

Notes: “Poverty only with medical expenses” is the share whose medical expenses push them into poverty (who would not otherwise be in poverty). “Poverty with or without medical expenses” is the share who are in poverty even without their medical expenses. Medical expenses are based on estimates of health insurance premiums and out-of-pocket health costs used to estimate the Supplemental Poverty Measure (SPM). The SPM is an alternative poverty measure published by the U.S. Census Bureau since 2010. In contrast to the Census Bureau’s “official” poverty threshold, which is benchmarked to three times the cost of a minimum food diet in 1963, the SPM takes into account the current cost of food, clothing, utilities, and shelter, as well as tax, work, medical, and child support expenses. The SPM also uses a broader resource measure that includes income and noncash benefits from both market sources and government programs (Fox and Burns 2021).

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of IPUMS Current Population Survey data (Flood et al. 2021).

Higher medical expenses explain much of the higher poverty experienced by older Americans, based on the Census Bureau’s Supplemental Poverty Measure (see chart note). About 2% of Americans ages 55–64 fall into poverty after paying their health insurance premiums and other out-of-pocket medical expenses. The effects of health expenses are more severe on older Americans, as many retirees live on near-poverty incomes. Nearly 3 in 10 poor seniors age 65 and older are poor because of medical expenses.

Despite near-universal Medicare coverage, medical expenses have a big impact on senior poverty because aging is associated with an increase in conditions requiring medical attention and Medicare does not cap out-of-pocket expenses. Younger Americans ages 25–54 are less likely to experience poverty as a result of medical expenses as they are more likely to be healthy, have sufficient income to keep them out of poverty, or postpone costly health care which, while reducing expenses temporarily, can impact their health and financial status at older ages (Montero et al. 2022).

Share this chart:

Copy the code below to embed this chart on your website.

Medical expenses drive many Black and Hispanic seniors into poverty: Poverty rates by age, race, and ethnicity, with and without medical expenses, 2021

3C
Poverty with or without medical expenses Poverty only with medical expenses Total poverty rate
White 5.4% 1.7% 0
Black 11.5% 1.6% 0
Hispanic 11.3% 2.4% 0
White 5.0% 2.4% 0
Black 12.1% 3.2% 0
Hispanic 13.9% 3.5% 0

 

ChartData Download data

The data below can be saved or copied directly into Excel.

Notes: “Poverty only with medical expenses” is the share whose medical expenses push them into poverty (who would not otherwise be in poverty). “Poverty with or without medical expenses” is the share who are in poverty regardless of their medical expenses. Medical expenses are based on estimates of health insurance premiums and out-of-pocket health costs used to estimate the Supplemental Poverty Measure (SPM). The SPM is an alternative poverty measure published by the U.S. Census Bureau since 2010.

Notes: “Poverty only with medical expenses” is the share whose medical expenses push them into poverty (who would not otherwise be in poverty). “Poverty with or without medical expenses” is the share who are in poverty regardless of their medical expenses. Medical expenses are based on estimates of health insurance premiums and out-of-pocket health costs used to estimate the Supplemental Poverty Measure (SPM). The SPM is an alternative poverty measure published by the U.S. Census Bureau since 2010.

In contrast to the Census Bureau’s “official” poverty threshold, which is benchmarked to three times the cost of a minimum food diet in 1963, the SPM takes into account the current cost of food, clothing, utilities, and shelter, as well as tax, work, medical, and child support expenses. The SPM also uses a broader resource measure that includes income and noncash benefits from both market sources and government programs (Fox and Burns 2021).

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of IPUMS Current Population Survey data (Flood et al. 2021).

Medical expenses contribute to very high poverty rates experienced by older Black and Hispanic Americans, based on the U.S. Census Bureau’s Supplemental Poverty Measure (see chart note). Black and Hispanic seniors age 65 and older are more than twice as likely to live in poverty as white seniors due to a combination of lower incomes and high medical expenses.

Among those ages 55–64, medical expenses have a larger impact on poverty for Hispanic than for white or Black Americans because Hispanic Americans are more likely to be uninsured (Keisler-Starkey and Bunch 2022). Black Americans are more likely to have near-poverty incomes than white Americans, but higher Medicaid eligibility limits the impact of health expenses on the share of Black Americans living in poverty because Medicaid caps out-of-pocket costs (Guth, Amula, and Hinton 2021).

Black and Hispanic Americans are more likely than white Americans to qualify for social insurance in the form of Medicare eligibility before age 65 (due to a long-term disability) and dual eligibility for Medicare and Medicaid (which covers some costs not covered by Medicare) (Ochieng et al. 2021). This helps to offset some—but not all—of the higher costs associated with poorer health.

Share this chart:

Copy the code below to embed this chart on your website.

Medicaid enrollment after age 65 is driven by long-term care needs: Percentage of adults who enroll in Medicaid after age 65, by number of years they receive long-term services and supports (LTSS)

3D
Percentage who enroll in Medicaid after age 65
All 29%
No LTSS 11%
Less than 2 years of LTSS 28%
2-4 years of LTSS 59%
5 or more years of LTSS 82%
ChartData Download data

The data below can be saved or copied directly into Excel.

Notes: Simulated results for adults born between 1941 and 1975. Long-term services and supports (LTSS), also referred to as long-term care, are health and social services for seniors and others whose age or health conditions limit their ability to care for themselves. LTSS include services provided in people’s homes, in community-based settings, and in nursing facilities. Estimates do not include unpaid care provided by family members and other caregivers.

Simulated results for adults born between 1941 and 1975. Long-term services and supports (LTSS), also referred to as long-term care, are health and social services for seniors and others whose age or health conditions limit their ability to care for themselves. LTSS include services provided in people’s homes, in community-based settings, and in nursing facilities. Estimates do not include unpaid care provided by family members and other caregivers.

Estimates are based on research by the Urban Institute for the U.S. Department of Health and Human Services using the DYNASIM4 microsimulation model, which starts with a representative population sample from the U.S. Census Bureau’s Survey of Income and Program Participation and calibrates income and health dynamics based on information from multiple surveys, including the Health and Retirement Study and the Medicare Current Beneficiary Survey.

Source: Johnson and Favreault (2020), Table 8.

Medicaid pays for nursing home care and other long-term services and supports (LTSS) for seniors with limited resources, including those who have drawn down their savings to pay for such care. High rates of Medicaid coverage therefore serve as a measure of the financial risks associated with the need for LTSS.

LTSS expenses are generally not covered by Medicare (Medicare.gov 2022). Private long-term care insurance, meanwhile, is often inaccessible and expensive while offering limited protection (Sammon 2020). Even among the small number of seniors with private insurance, about a quarter will let their policies lapse, often due to cognitive impairments that make them more likely to need the long-term care that the insurance would have paid for (Friedberg et al. 2017).

While only 11% of seniors without long-term care needs enroll in Medicaid after age 65, most seniors who require long-term care for two or more years end up in the means-tested program. An estimated 59% of seniors requiring two to four years of LTSS, and 82% of those requiring five or more years of LTSS, will end up on Medicaid. Nursing home care is particularly expensive, with 77% of seniors who require nursing home care for two or more years enrolling in Medicaid (Johnson and Favreault 2020; not shown in chart).

Share this chart:

Copy the code below to embed this chart on your website.

Low-income seniors are most likely to need long-term care, but all seniors are at risk: Share of adults receiving two or more years of long-term services and supports after age 65, by lifetime earnings quintile

3E
Lifetime earnings quintile  2 years or more (%)
Top 20% 21%
Fourth 20%  20%
Middle 20% 20%
Second 20% 23%
Bottom 20% 31%
All age 65+ adults 23%
ChartData Download data

The data below can be saved or copied directly into Excel.

Notes: Long-term services and supports (LTSS), also referred to as long-term care, are health and social services for seniors and others whose age or health conditions limit their ability to care for themselves. LTSS include services provided in people’s homes, in community-based settings, and in nursing facilities. Estimates do not include unpaid care provided by family members and other caregivers.

Notes: Long-term services and supports (LTSS), also referred to as long-term care, are health and social services for seniors and others whose age or health conditions limit their ability to care for themselves. LTSS include services provided in people’s homes, in community-based settings, and in nursing facilities. Estimates do not include unpaid care provided by family members and other caregivers.

Estimates are based on research by the Urban Institute for the U.S. Department of Health and Human Services using the DYNASIM4 microsimulation model, which starts with a representative population sample from the U.S. Census Bureau’s Survey of Income and Program Participation and calibrates income and health dynamics based on information from multiple surveys, including the Health and Retirement Study and the Medicare Current Beneficiary Survey.

Source: Johnson and Favreault (2020), Table 5.

Low earners face a greater risk than higher earners of requiring long-term care due to well-documented disparities in health and disability by socioeconomic status (Isaacs et al. 2021). Among seniors age 65 and older in the lowest lifetime earnings fifth, more than 3 in 10 (31%) will require two or more years of long-term services and supports (LTSS). Among seniors in the four higher earnings quintiles, roughly 2 in 10 will need two-plus years of LTSS.

Note that there is little difference in risk of needing LTSS among the latter four groups. This is likely because higher earners are more likely to live long enough to develop health conditions associated with advanced old age, offsetting their other health advantages relative to lower earners (Johnson 2019).). Unlike individuals in the top four lifetime earnings groups, however, the adverse health effects of living in or near poverty for the bottom fifth of lifetime earners are not offset by their shorter lifespans and they face a higher risk of LTSS needs.

Though low earners are most affected, all seniors face a significant risk of needing to pay for LTSS for two or more years, with a concomitant increased likelihood of needing Medicaid to help with costs (see Chart 3D). Seniors 65 and older who need expensive nursing home care are at especially high risk of exhausting their resources. Even among those in the top earnings quintile, 43% of those who need nursing home care for two or more years end up on Medicaid (Johnson 2019; not shown in chart).

Share this chart:

Copy the code below to embed this chart on your website.

Low-income older households had little capacity to cope with financial shocks even before the COVID-19 pandemic and recession: Share of working households ages 55–64 that are financially fragile, by income group, 1992–2018

3F
Year Bottom 50% 50th–90th percentile Top 10%
1992 35.0% 20.5% 13.8%
1994 33.3% 17.7% 15.2%
1996 36.0% 19.5% 15.0%
1998 37.9% 20.0% 11.1%
2000 37.9% 22.1% 11.0%
2002 36.3% 21.1% 14.7%
2004 41.1% 25.4% 15.2%
2006 44.6% 24.2% 10.8%
2008 44.6% 31.0% 16.4%
2010 51.1% 34.1% 18.0%
2012 49.1% 35.6% 19.5%
2014 51.0% 32.3% 16.9%
2016 50.1% 31.0% 15.3%
2018 54.2% 29.6% 15.6%
ChartData Download data

The data below can be saved or copied directly into Excel.

Notes: A household is deemed financially fragile if it exceeds at least one of four thresholds: a home mortgage loan-to-value ratio above 80%; a ratio of nonhousing debt to liquid assets above 50%; less than three months’ worth of income in liquid assets; or rent exceeding 30% of income. Sample includes households with at least one working member and one member age 55–64. For married and partnered households, income percentiles are determined based on total household income divided by 1.7 to account for the fact that living expenses for couples are higher than—but less than double—the expenses of single householders.

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of Health and Retirement Study (HRS) microdata (RAND and University of Michigan 1992–2018).

Many older working American households were struggling financially before COVID hit. These households therefore had less of a financial cushion to protect them from the economic fallout of the pandemic.

Over half (54%) of lower-income (bottom 50%) households ages 55–64 were financially fragile before the COVID-19 pandemic, based on their debt burdens, housing costs, and the savings they had available to access in an emergency. This is an increase in financial fragility from a third (35%) of such households in 1992.

In the wake of the Great Recession, rising mortgage debt, credit card balances, auto loans, and student debt hurt older households’ finances (GAO 2021; Butrica and Mudrazija 2020). As seen in the chart, older households across the income distribution experienced growing financial fragility between 2008 and 2012.

Wealthier older households—those in the top 10% of the income distribution—have roughly recovered to their 2008 levels. Households in the 50th–90th percentiles have also recovered to their 2008 levels but still face much higher rates of financial fragility than those in the top 10%; they are also much more financially fragile than their counterparts were in 1992. Households in the bottom half of the income distribution did not recover well from the Great Recession, and by 2018 they had reached historically high rates of financial fragility.

Other studies have found similar trends. A recent study by the Government Accountability Office (GAO) found that older households are more likely to be indebted than they were three decades ago, and a typical older household age 50 or older held roughly three times as much debt in 2016 as it did in 1989, adjusted for inflation (GAO 2021, not shown in chart). Butrica and Mudrazija (2020) found a significant increase in debt and falling credit score, a sign of deteriorating financial stability, among households age 70 and older, mostly due to increases in mortgage debt.

Rising debt levels are not necessarily cause for concern if they reflect rising homeownership or access to higher education among older households. However, a closer look at trends in indebtedness—such as rising home mortgage loan-to-value ratios among older homeowners—cautions against such a rosy view.

Share this chart:

Copy the code below to embed this chart on your website.

Older Black and Hispanic working households were stretched thin before the COVID-19 pandemic and recession hit: Share of working households ages 55–64 that are financially fragile, by race and ethnicity, 1992–2018

3G
Year White Black Hispanic
1992 22.9% 43.3% 43.0%
1994 20.7% 41.8% 38.2%
1996 22.7% 40.4% 39.9%
1998 22.8% 43.0% 43.8%
2000 23.5% 46.7% 43.8%
2002 23.1% 45.6% 38.5%
2004 27.6% 49.9% 38.4%
2006 27.0% 47.8% 44.6%
2008 32.1% 54.2% 39.4%
2010 34.9% 56.5% 49.4%
2012 35.7% 54.5% 46.0%
2014 34.3% 51.8% 49.9%
2016 33.6% 54.1% 45.5%
2018 33.4% 57.0% 50.7%

 

ChartData Download data

The data below can be saved or copied directly into Excel.

Notes: A household is deemed financially fragile if it exceeds at least one of four thresholds: a home mortgage loan-to-value ratio above 80%; a ratio of nonhousing debt to liquid assets above 50%; less than three months’ worth of income in liquid assets; or rent exceeding 30% of income. Sample includes households with at least one working member and one member age 55–64. For married and partnered households, income percentiles are determined based on total household income divided by 1.7 to account for the fact that living expenses for couples are higher than—but less than double—the expenses of single householders.

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of Health and Retirement Study (HRS) microdata (RAND and University of Michigan 1992–2018).

Among working households ages 55–64, over half of Black (57.0%) and Hispanic (50.7%) households were financially fragile before the COVID-19 pandemic, based on their debt burdens, housing costs, and savings they could access in an emergency. In contrast, only a third (33.4%) of older white households were stretched too thin to weather a financial shock.

These findings are in line with previous research showing that debt burdens have risen more for Black and Hispanic households than for white households. Between 1989 and 2016, the debt-to-asset ratio of the typical household age 50 or older increased from 8% to 17% for white households, from 16% to 35% for Black households, and from 17% to 37% for Hispanic households (GAO 2021).

Share this chart:

Copy the code below to embed this chart on your website.

Older Black households are much more likely than white or Hispanic households to have education debt: Percentage of households ages 55–64 with education loan debt, by race and ethnicity, 1992 and 2019

3H
Share of households with education loan debt White Black Hispanic
1992 3.2% 3.6% 0.2%
2019 11.5% 18.4% 7.3%
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: In the Survey of Consumer Finances, age and other household characteristics are based on the reference person, defined as the individual for a single householder, the male in a mixed-sex couple, and the older person in a same-sex couple.

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of Survey of Consumer Finances microdata (Federal Reserve 1992 and 2019).

Student loan debt is the fastest-growing type of debt among older American households. More than four times as many households ages 55–64 had student loan debt in 2019 compared with 30 years ago (12.2% in 2019 vs. 2.9% in 1992; not shown in the chart). At the same time, their debt burden has increased: The median education debt-to-earnings ratio (total student loan debt to annual earnings) almost doubled, from 15.8% in 1992 to 28.4% in 2019 (Schuster 2021).

Ballooning student loan debt among households ages 55–64 is only partly explained by an increase in the share of older Americans with bachelor’s degrees, which rose from 17.9% to 32.4% over this period (authors’ analysis of Current Population Survey microdata [Flood et al. 2021]; not shown in the chart). Some of the increase in older households’ student debt reflects parents borrowing to help pay for their children’s educations (Looney and Lee 2018). College costs have risen rapidly (Jackson and Saenz 2021), while many Americans are burdened with student loans despite not obtaining—or their children not obtaining—degrees (Siegel Bernard 2022).

Black households have seen the fastest increase in student loan debt. As shown in the chart, the share of Black households ages 55–64 with student loan debt grew fivefold between 1992 and 2019, while the share of Black Americans in this age group with bachelor’s degrees roughly doubled, from 8.1% to 17.9% (authors’ analysis of Current Population Survey microdata [Flood et al. 2021]; not shown in chart). This suggests that many of these households took on student loan debt for their children or grandchildren, or that the student loans are for their own education but college costs rose faster than earnings and fewer former students were able to pay off their debts before age 55. In either case, the higher debt puts additional pressure on these older households’ finances.

Share this chart:

Copy the code below to embed this chart on your website.

Older workers experienced higher unemployment than mid-career workers in the pandemic recession: Percentage-point difference in average unemployment rate between older workers (age 55+) and mid-career workers (ages 35–54), 1975–2020 recessions

3I
Months of peak unemployment Gap between mid-career (35–54) and older workers (55+)
May–Oct 1975 -0.8
Jul–Dec 1980 -1.2
Dec 1982–May 1983 -1.7
Jun–Nov 1992 -0.6
Jun–Nov 2003 -0.5
Oct 2009–Mar 2010 -1.2
Apr–Sep 2020 1.1

 

ChartData Download data

The data below can be saved or copied directly into Excel.

Notes: Chart shows the percentage-point difference (how much higher or lower the average unemployment rate of older workers age 55+ is relative to that of mid-career workers ages 35–54) over six-month periods beginning in the month of peak unemployment in each recession.

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of Current Population Survey microdata (Flood et al. 2021).

The recession triggered by the COVID-19 pandemic was highly unusual in that older workers suffered greater job losses than mid-career workers. In the six-month period from April to September 2020, the unemployment rate for workers age 55 and older averaged 9.7% (not shown in the chart), more than a percentage point higher (+1.1 ppts.) than the 8.6% unemployment rate for workers ages 35–54. In contrast, from October 2009 to May 2010, the peak unemployment months of the Great Recession, the unemployment rate of older workers averaged 7.0%, 1.2 percentage points lower than the 8.2% unemployment rate of mid-career workers.

In typical recessions, older workers are less likely to be laid off than mid-career workers because they usually have more work experience and seniority. These factors offered less protection during the pandemic, since pandemic job losses were driven by the mass shutdown of nonessential sectors and a shift in consumer spending from services to goods.

Share this chart:

Copy the code below to embed this chart on your website.

Pandemic job losses among workers 65 and older were steep and persistent: Percentage changes in employment rates from pre-pandemic peaks, by age group, February 2020–August 2022

3J
Year Ages 35–54 Ages 55–64 Age 65+
Feb-2020 0.0% 0.0% 0.0%
Mar-2020 -0.7% -0.7% -4.5%
Apr-2020 -11.8% -11.8% -18.6%
May-2020 -9.6% -9.6% -18.1%
Jun-2020 -6.9% -7.7% -14.9%
Jul-2020 -6.6% -7.2% -12.0%
Aug-2020 -4.9% -5.7% -10.6%
Sep-2020 -5.1% -5.1% -7.6%
Oct-2020 -4.2% -4.0% -6.6%
Nov-2020 -4.0% -4.9% -6.3%
Dec-2020 -4.3% -5.2% -8.8%
Jan-2021 -4.6% -5.9% -11.1%
Feb-2021 -4.6% -4.2% -12.8%
Mar-2021 -4.1% -3.3% -11.4%
Apr-2021 -4.2% -3.3% -9.9%
May-2021 -4.0% -2.2% -10.3%
Jun-2021 -4.1% -3.5% -11.8%
Jul-2021 -3.3% -3.6% -12.1%
Aug-2021 -2.8% -3.1% -8.9%
Sep-2021 -2.0% -1.9% -6.3%
Oct-2021 -1.8% -2.1% -7.0%
Nov-2021 -1.2% -2.2% -6.0%
Dec-2021 -1.1% -1.9% -5.3%
Jan-2022 -1.5% -1.8% -7.7%
Feb-2022 -0.9% -1.1% -5.9%
Mar-2022 -0.1% -0.1% -7.0%
Apr-2022 -0.4% -0.4% -7.4%
May-2022 -0.3% -0.2% -6.1%
Jun-2022 -1.0% -1.5% -7.8%
Jul-2022 -1.0% -1.0% -7.4%
Aug-2022 -0.3% -1.7% -5.8%
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Chart shows percentage changes in employment-to-population ratios relative to February 2020, the peak month of economic activity before the pandemic recession.

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of Current Population Survey microdata (Flood et al. 2021).

The employment rate of seniors age 65 and older plummeted 18.6% between February and April 2020, an even steeper drop than that experienced by mid-career workers ages 35–54 and older workers ages 55–64, who both experienced 11.8% declines in employment rates.

Employment losses among workers age 65 and older were also more persistent than those of younger workers. The employment rate of mid-career workers had essentially recovered by August 2022; that of older workers ages 55–64 was only slightly below its pre-pandemic level (-1.7%). However, the employment rate of seniors remained -5.8% below its pre-pandemic level two and a half years later.

Some of the employment decline among older workers reflected an increase in retirements. Older workers in part-time jobs, or in occupations characterized by high physical proximity to other workers or to customers, were especially likely to call it quits (Davis 2021). For these workers, COVID disruptions, health and safety concerns, caregiving responsibilities, rising net worth, or other factors tipped the balance in favor of retirement. But for the majority of older workers, leaving the labor force was triggered by job loss (Davis and Radpour 2021).

Most unemployed older workers returned to the workforce, aided by a rapid recovery brought about by stimulus checks, expanded unemployment insurance, and other timely countercyclical measures enacted by Congress (CBPP 2022). Though the pandemic recession had an unusually severe impact on older workers, these workers would likely have fared much worse in a slower recovery with fewer supports for unemployed workers and their families, especially since unemployed workers in their 50s and older tend to remain out of work longer than young or mid-career workers (Johnson and Butrica 2012; Johnson and Gosselin 2018).

Share this chart:

Copy the code below to embed this chart on your website.

Non-college-educated older workers saw greater job losses than their college-educated counterparts during the pandemic recession: Percentage changes in employment rates by age and education, February–April 2020

3K
Employment percent change
No college degree -13.9%
Bachelor’s degree or more -9.2%
No college degree -21.0%
Bachelor’s degree or more -15.6%
ChartData Download data

The data below can be saved or copied directly into Excel.

Note: Chart shows the percentage change in the employment-to-population ratio from February 2020, the month before the pandemic recession, to April 2020, the trough of the recession.

Source: Economic Policy Institute (EPI) and Schwartz Center for Economic Policy Analysis (SCEPA) analysis of Current Population Survey microdata (Flood et al. 2021).

When the pandemic recession hit in early 2020, older workers without bachelor’s degrees experienced greater job losses than their college-educated counterparts. The most affected were workers over 65 without college degrees, 1 in 5 (21.0%) of whom found themselves out of work.

While many white-collar workers have been able to work from home during the pandemic, most noncollege workers in service-sector jobs did not have that option (Gould 2020). Many of these workers were laid off or furloughed. Those who remained in the workforce were more likely to be exposed to COVID-19 health risks. These risks were particularly acute for older workers in meatpacking, caregiving, and other low-paid service jobs often deemed “essential” but not adequately compensated or protected (Hassan 2021; Farmand et al. 2020; Lewis 2021).

Though job losses were highest among workers 65 and older and among older workers without college degrees, some professional occupations, including teachers and nurses, have seen waves of early retirements due to deteriorating working conditions during the pandemic (Barnes 2022; Zhavoronkova et al. 2022).

Share this chart:

Copy the code below to embed this chart on your website.


See more work by Monique Morrissey, Siavash Radpour, and Barbara Schuster